Zone System


I have been blogging about various aspects of the Zone System here for some time now. That means that there is some useful stuff scattered about the site, and you, the reader, have to go hunting for it. I’ve decided to collect this material in a set of pages dedicated to aspects of the Zone System in a way that I hope makes sense to modern photographers. I have met many photographers who are are inherently non-mathematical. Let me rephrase that, almost every photographer I ever met is inherently non-mathematical. Somehow, the inverse square law for the fall-off of light defeats them, in spite of they’re being visual beings. Turning the inverse square law into stops of light versus distance becomes a matter of real pain. Therefore, I am going to try to be as gentle as possible with the hard math, and even with the intermediate math. IMHO, there is no point in mathematics without understanding. Where possible, I’ll use graphs, charts, diagrams, and illustrative images.

A serious question to consider before going any further is, why? Why bother with the Zone System in these days of digital? It is apparent that there are many excellent digital images that have been and are being produced, and almost no one uses the Zone System even if they are aware of it. If you have read this far, you are more than likely aware of the images of Ansel Adams, one of the fathers of the Zone System. His iconic and superb B&W images, whether landscapes or portraits, have been the motivation for many of us to try our hands at photography. The visual impact of their large swathes of deep black and soft floating whites never fails to evoke a sense of beauty that equals or exceeds that of the original scene. These B&W images are, obviously, far from the original scenes in any case, which would have been viewed by any human observer in color. While Adams’ images were derived from the original scenes, they are, in no way, accurate representations of them. How could they be? Daytime skies are not black. Trees and faces are not gray. Billowing clouds maybe nearly white, but they are not the paper white of a print.

Instead, Adams’ images are almost the perfect visual demonstration of the balance of yin and yang, light and dark, of Taoist philosophy. To achieve this balance, Adams employed the Zone System to take whatever the range of light was in a given scene and create negative and print that presented the viewer with his signature deep blacks and delicate whites. This harmony is not something that just naturally exists in a B&W image created by any of the available conversion techniques that you may have encountered, from desaturating a color file to using products like Nik’s Silver Efex or Grubba Software’s TrueGrain; but a proper application of the Zone System can help.

The Zone System is inherently an experimental and pragmatic approach to achieving this yin and yang balance in a B&W image. What I believe has been lost in any and all of the current “digital” approaches to the Zone System is that they are arbitrary and logical systems concerning the properties of digital files without any reference to their color spaces, tone curves, or the cameras, sensors, and exposures from which they were derived. In the following exposition, I hope to correct this problem and demonstrate, I hope, something of the spirit in which Adams himself might have tackled the problem of revising his Zone System for modern digital cameras, post processing, and printing.

A simple idea

What is going to be different in my presentation than those of other writers on this topic, at least in the modern digital context, is that I will be basing my work on actual measurements. “Zones” are often described in terms of where an author decides they should be, or wants them to be, or thinks that they should be. In the following, Zones are measured. The measurements are mine or those of others. Where they are others, I tell you who made them and where to follow up. This is the origin of Zones in Adams’ say, based on a science called sensitometry. Ansel Adams described the Zone System as a practical application of sensitometry. In the days of film and dark rooms, sensitometry related the density of silver on negatives and prints to their exposure to light. The digital photographer can still measure the density of ink on a print, but the negative has disappeared from the digital workflow, and a replacement for density must be sought. Alternatively, a deeper understanding of what density means in the digital context must be achieved. In these pages, I offer an easy and available alternative. The L-channel of the LAB color space ties together density (D), grayscale ink measure (K), and ratios of reflected and transmitted light (Y/Y_n) in ways that make sense in terms of human visual perception of scenes, prints, and images rendered on screens.

In its heart of hearts, the Zone System is a truly simple idea. It is a scale that concerns the intensity of light. Ansel Adams would say, the “range of light”, which was cute, since he’d be shooting mountain ranges and that made a nice pun. In another way, it also makes a great deal of sense since some scenes have a much greater range of light than others. This is a point that I’ll have to return to in what follows.

Allow me to begin with this simple idea of the range of light of this medium:

Zones for the monitor

Zones for the monitor

Here we have the Zone System in its ultimate simplicity, as an expression of the range of light that our monitors can handle. There is a pure white square at 100%, a pure black square at 0%, and a middle gray square at 50%. Or at least, that’s the idea. Getting the values of 100% and 0% from your monitor seems an easy enough thing to do, at least superficially. It is actually much more complex than you might give it credit for. But getting 50% turns out to be very very complex indeed. First of all, 50% of what? If it’s 50% of whatever is right for my monitor, what is that on yours? Or every other monitor that might ever render this page? It is the struggle with the meaning of 50% that is the essence of the Zone System. More of this later.

Sources of light and surfaces

In most scenes that photographers have to deal with, there are no direct sources of light. I will admit that this is not always so: sometimes we’ll shoot a scene with a candle, a street lamp, stars, a headlight, or even, heaven help us, the sun. But hold that thought for a moment, because they’ll turn out to be less unusual than you might imagine. Back to indirect light… This means that our scenes are comprised of surfaces that are either scattering or reflecting light. In photography jargon, we say that these surfaces are either diffuse or specular. In everyday English, we use words like matte, flat, dull, drab, muted, lusterless, and so on, to describe a surface that scatters incident light in all directions. It might be more correct to say that the light that comes off such a surface is diffuse, but photography jargon calls the surface itself diffuse. In contrast, there are the highly reflective surfaces of mirrors, metal, and to some extent, water and glass. Mirrors and metal will reflect incident light at any angle. Other surfaces, like glass and water, will reflect light only if the angle is greater than the Bragg angle. We have all seen this effect, even if the physics behind it is difficult. If we look straight through a pane of glass, it is transparent. Even if there is a light right behind us. However, shift our point of view somewhat so that the light on our side of the glass is at some angle, and we see its internal reflection from the surface of the glass. To describe specular surfaces, we use words like reflective, mirrored, mirror-like, shiny, sheen, glossy, metallic, and so on. However, there are other sorts of specular objects that lack the uniform quality of a mirror or metal surface; for example, clouds, which are huge volumes filled with water droplets, each one of which is a tiny reflective element. In some ways, these droplets scatter light diffusely; in other ways, they are mirror-like.

Diffuse and specular

So, we have these two concepts: diffuse and specular. No real surface is purely one or the other, although they can get “close enough for government work”. For a photographer, a real mirror is pretty specular; and a wall painted with a dark, flat, matte paint is pretty diffuse. The skin of a subject’s face is somewhere in between, with an ability to show sheen under direct flash or sunlight. And so on.

Working back from the end

Photography is about capturing light from a real scene and producing an image of that scene for viewing in some other medium. The classic medium has been on a paper surface, but now I’d wager that some form of screen is much more typical. We can also print on canvas, metal, cloth, acrylic, plastics, and who knows what else. The Zone System was developed at a time when a photographic print was almost always delivered on paper. It was natural to begin the struggle with what 0% – 50% – 100% meant in the context of paper, and more specifically, in the context of B&W photographic paper. At the present, we must also struggle with what 0% – 50% – 100% means in the context of computer monitors, television sets, tablets, smartphones, and the entire raft of digital devices that could be used to render and display photographic images.

The capture medium

From the final medium, the struggle leads backwards to the medium that stores the image exposure; that is, to a negative, or more currently, to a digital file or files. We are faced with the meaning of 0% – 50% -100% in the context of a negative or a digital file. We shall find that this has a fairly straightforward interpretation in the context of a negative, once we have solved the problem for a print. However, the case of digital representations of scene exposures is anything but straightforward. There is an additional subtlety in the context of a digital file that does not arise in the context of a negative. A negative is a real physical object. We can look at it in the same way that we can look at a print. Whatever ‘50% of the incident light reflecting back from a print’ means has a similar interpretation in ‘50% of the incident light being transmitted through a negative’. But a digital file doesn’t do anything to light. It is some abstract mathematical representation of light, and it turns out that there are a wide variety of methods to turn a digital image file into either a print or an image on a screen, or into another kind of file. We have NEFs and JPGs and PNGs and TIFFs and PSDs and each one may or may not allow for variations in color space and tone curve which may or may not be honored by different rendering methods used by different applications on different computing platforms (combinations of hardware and operating system).

Digital media

Let me make this a little more clear (perhaps as clear as mud). Here are screen shots of two “hex dumps” of image files on my Mac. (For those in the know, they are dumps of the data forks.) They are the first few lines of the file, not the entire thing in either case. That would go on for screen after screen after screen. The first is for the raw NEF and the second for a size-reduced JPG.

Hex dump of NEF

Hex dump of NEF

Hex dump of JPG

Hex dump of JPG

Each pair of characters in the body of the table represents 8 bits of data. Each character can be on of the 16 elements {0, 1, 2, 3, 4, 5, 6, 7, 8, 9, A, B, C, D, E, F}. These are the so-called “hexadecimal” digits, used to count in base 16. They are used because they are a reasonably compact way to represent four binary digits; for example, “F” means “1111”, “A” means “1010” and so on. If all of this is Greek to you, fear not. My first point is just that these image files are bags of bits. My second point is that they are different bags of bits. Yet, you would be hard pressed to distinguish between the thumbnails for these two files as rendered on either my computer, or yours.

Where is the middle?

You may be aware that the maximum value that can be expressed with 8 binary digits is, as a decimal number, 255. The actual count runs from 0 (decimal) to 255 (decimal) with 256 elements. A standard JPG file encodes each pixel as three numbers for each color element; that is, for red (R), green (G), and blue (B). With this knowledge in hand, it might seem that 50% would be an RGB value of (128, 128, 128). Perhaps mysteriously, this is not true. For a standard JPG file, 50% gray would be more like (119, 119, 119).

So, we have another struggle, as photographers, to come to an understanding of what 0% – 50% – 100% means for the digital files that we employ and edit every day. We’ll find that for files in an sRGB or AdobeRGB color space, it’s 0 – 119 – 255. For a file in a ProPhotoRGB color space, it’s 0 – 100 – 255. If this does not instantly strike you as odd, something is wrong.

If you have a Mac computer, you will have an application in your Applications/Utilities folder called DigitalColorMeter. I invite you to find that and open it up. Scroll back to the image above showing the white, gray, and black squares. Set the application to read “native values” and read out what you get for the gray square. Here is what shows up on my monitor:

Measuring the native RGB of gray

Measuring the native RGB of gray

And yet, this is what shows up when I look at the original file in Photoshop:

Measuring the original file in Photoshop

Measuring the original file in Photoshop

But this is what Photoshop tells me the values are:

What Photoshop says

What Photoshop says

Really, where is the middle!?!

Let us make things even stranger. Here are two screen shots of just the gray square, one taken from the screen display of Photoshop on my Mac and the other from the screen display of the JPG that Photoshop has exported as rendered by the Safari browser on my Mac.

The difference is subtle, but there. I invite you to use the DigitalColorMeter to check. The Photoshop screen shot is lighter. The difference has much to do with how my monitor is calibrated, how exporting for the web works in Photoshop, and how I’ve done the screen shots. My point in this brief exercise is that, even on a single platform, what 50% means on the screen is ambiguous and confusing at best. If this is frustrating, join the club. For all its benefits, digital introduces complexities that seem to defy the imagination.

I shall aim in what follows to clear up these issues and explain what just happened in a way that makes sense. But it should be apparent that the notion of what 0% – 50% – 100% means for digital is going to be some trouble to straighten out. Don’t worry. It will make sense in the end.

The original scene

As we work our way backwards, we then encounter the original scene and its own range of light. We, again, have to decide what 0% – 50% – 100% means in this context. It will turn out, once more, that what 50% meant in the context of the print and the negative is equally useful in the context of the scene. However, to our apparent surprise, the range of light will often involve values of greater than 100%. While such values make little sense in the context of a print, a screen, or a negative, they make perfect sense in the context of a real scene. Read on, and you will first see how this can be, and what can be done about it.

The exposure scale

In part, this is because, if the Zone System is a measurement scale, then its units turn out to be rather odd. In fact, contrary to what you often read in second sources, Adams himself made a distinction between “Zones” and “Values”. A Value has a particular meaning in this context; but a Zone is a matter of choice. In other words, a photographer measures a Value, but selects a Zone. This is to be done intentionally at the time the scene is photographed. Following Adams’ usage, I should distinguish between the Value that some element in a scene might have and the Zone that you or I as a photographer would map this too. We set this relationship by measuring the Value of some reference element in the scene (perhaps a gray card) and then choosing an exposure setting. In Zone System jargon, it is customary to refer to whatever the luminous intensity of a middle gray card is in a scene as “Value V”. Hence, Value V depends not just on the scene, but on the light. Value V for a landscape at noon is not Value V for the same landscape at 7pm. It is not the same at noon on a sunny day and at noon on the following, cloudy day. To quote Adams, “a middle-gray print value that matches the 18 percent reflectance gray card is designated Value V.” (From The Negative, p.48.) The simple choice that a photographer can make is to map Value V to Zone V, but there are so many times that this is the wrong choice, and a stunningly wrong choice at that, that we shall have to consider very clearly the entire chain from scene to digital file (or files) to print (or screen). We shall discover that even if we ultimately want Value V in our scene to map to Value V in our final print, as is very often the case, that choosing an exposure that puts Value V on Zone V is a big mistake, especially with a DSLR. And yet, this is the default behavior of most cameras (or close to it).

We just learned from Adams about an “18 percent reflectance gray card”, and it will turn out that this will become our reference for what 50% means in terms of prints, negatives, and scenes. You will learn that this rather odd choice of mid-point (I mean after all, what is 18% the middle of? 36%?) comes from studies of human visual perception. It might perhaps surprise you to learn that the middle gray value in the square we’ve been looking at on this page is probably putting out about 22% of the maximum amount of light that your monitor can emit; that is, its luminous intensity is around 22% of what the white square is showing. I base my guess on the belief that most monitors are set up using an sRGB color space or something close to it. So here is another quandary that we shall have to resolve: what does 22% have to do with 50%? Back to Adams, what does 18% have to do with 50% for that matter? And then, if 18% is right, can 22% be right too? (Hang on, but if you’re a ProPhotoRGB kind of a person, for you 50% is 29%. Yet another quandary!)

The units of the scale

Still, we have just been battling with the question of a single useful reference point on the range of light for prints, screens, files, and scenes. We have yet to introduce a unit of measure, a scale. Following the standard conventions of photography and the Zone System, our scale will be the “stop”. A one-stop change in luminous intensity is, depending on the direction, a doubling or halving of the amount of light. (If you are familiar with sound, it is equivalent to a 3dB change in sound level.) We get a one-stop change of exposure by doubling or halving the ISO value, by doubling or halving the shutter speed, or by changing the aperture setting by the square root of 2.

The pieces of the puzzle

So here are the guts of the Zone System: We have a scale for the range of light. The scale has a reference point at the middle, call it 50% for lack of a better term. The units of the scale are stops. The scale reads off in Values. The reference point in the middle is Value V, by definition. The scale of Values goes up and down from Value V by stops; so one stop more than V is VI, one stop less is IV, and so on. When we make an exposure, we choose what Value from the light in the scene to map to what Zone. A ‘normal’ exposure might map Value V to Zone V. An underexposure might map Value V to Zone IV.

In employing the Zone System, this is our first and central degree of freedom: we choose an exposure setting. In making this choice, we are, intentionally or unintentionally, mapping a Value V luminous intensity to some exposure Zone. If, for the sake of argument or experiment, I actually do put a gray card in a scene and take a picture with my DSLR, the values in the file can turn out to be greater, lesser, or exactly on whatever 50% means. If I shoot JPG, that would be 119 decimal. If I shoot 12-bit NEF, that would be about 737 decimal. Our next degree of freedom is associated with how we “develop” the negative. In the traditional dark room, this meant using either normal development, or pushing (over-developing) or pulling (under-developing) the development by some number of stops. In a digital dark room, the equivalent effects can be achieved through a curves layer. You will learn that a curves layer is a much more powerful and flexible tool than push or pull dark room processing could ever hope to be.

As Alice once said, “what is the use of a book without pictures or conversations?” So, down the rabbit hole we go, looking for the pictures and the conversations!

PS. This is a work in progress. Stay tuned for updates and new pages as I add them in.

Next Page…